Mineral bridges in nacre revisited

Antonio G. Checa,1 Julyan H. E. Cartwright,2 Marc-Georg Willinger3,4

1
Departamento de Estratigrafa y Paleontologa, Facultad de Ciencias, Universidad de Granada, E-18071

Granada, Spain. 2Instituto Andaluz de Ciencias de la Tierra, CSICUniversidad de Granada, Campus

Fuentenueva, E-18071 Granada, Spain. 3Laboratorio Associado CICECO, Campus Universitario de Santiago,

Universidade de Aveiro, 3810-193 Aveiro, Portugal. 4Department of Inorganic Chemistry, Fritz-Haber-

Institute of the Max-Planck-Society, Faradayweg 4-6, 14195 Berlin, Germany.

Abstract

We confirm with high-resolution techniques the existence of mineral bridges between superposed

nacre tablets. In the towered nacre of both gastropods and the cephalopod Nautilus there are large

bridges aligned along the tower axes, corresponding to gaps (150-200 nm) in the interlamellar

membranes. Gaps are produced by the interaction of the nascent tablets with a surface membrane

that covers the nacre compartment. In the terraced nacre of bivalves bridges associated with

elongated gaps in the interlamellar membrane (> 100 nm) have mainly been found at or close to the

edges of superposed parental tablets. To explain this placement, we hypothesize that the

interlamellar membrane breaks due to differences in osmotic pressure across it when the

interlamellar space below becomes reduced at an advanced stage of calcification. In no cases are the

minor connections between superimposed tablets (< 60 nm), earlier reported to be mineral bridges,

found to be such.

1"

"
Introduction

Nacre is a lamellar biomaterial present today in representatives with old affinities of the molluscan

classes Bivalvia, Cephalopoda, Gastropoda and Monoplacophora. It has a brick and mortar

structure, in which the bricks are aragonite tablets (5-15 m wide) and the mortar is composed of

organic material (proteins and polysaccharides). Its secretion is fully extracellular and begins with

the production of very closely spaced (~100 nm) parallel interlamellar membranes (Nakahara, 1983,

1991) through liquid crystallization (Cartwright and Checa, 2007, Cartwright et al., 2009). These

consist of a core of -chitin that is surrounded by acidic proteins (Levi-Kalisman et al., 2001). The

intermediate spaces are filled with a gel enriched in silk fibroin. Nacre tablets begin to grow within

these intermediate spaces, first quickly along the vertical direction (coincident with the aragonite c-

axis), at the same time as the interlamellar membranes separate to the required distance to

accommodate the final thickness of the aragonite plates. Tablets later expand sidewards until

impinging upon their neighbours, so as to fill all of the available volume.

One of the main current debates concerning nacre construction is whether the superimposed

tablets are continuous or not across so called mineral bridges (MBs). This debate emerged with

the description of nanopores (5-50 nm in diameter) in the decalcified interlamellar membranes of

the gastropod Haliotis rufescens and the proposal that a new tablet would initiate as an offshoot of

an underlying tablet through the nanopore, with no need for a new nucleation event each time a

tablet is formed (Schffer et al., 1997). The mineral bridge hypothesis provided an interesting

alternative to the previous heteroepitaxial hypothesis of nucleation of nacre tablets directly onto the

organic scaffolding (Weiner and Traub, 1980, 1981).

2"

"
Fig. 1. Arrangement of nacre in molluscs. (A) Towered nacre in the gastropod Osilinus lineatus. (B) Marginal

nacre of the internal shell surface of the cephalopod Nautilus pompilius. (C) Septal nacre of Nautilus

pompilius. The distribution is tower-like, with up to four piled up tablets. (D) Marginal nacre of the bivalve

Nucula sulcata. Tablets grow in low towers. (E) Internal nacre of Nucula sulcata. Newborn tablets always

appear at the margin of parental tablets. There are two to null tablets per parental tablet. (F) Nacre of Gibbula

umbilicalis. Note the general agreement in orientation between the lineations of parental and filial tablets.

Some new crystals may appear (arrows) or disappear in the newborn tablets. According to lineations,

polycrystalline (twinned) tablets are dominant. (G) Etched nacre of Pinctada margaritifera. The lineations

revealed upon etching coincide with the crystallographic a-axis. Lineations of filial and parental tablets are

parallel. (H) Diffraction contrast STEM view of Gibbula umbilicalis nacre. Similar greyscale intensities

indicate even crystallographic alignment of tablets. (I) Diffraction contrast TEM photograph of Pinctada

margaritifera. Evenly oriented tablets are arranged in a stair-like fashion. Growth direction is to the top left.

3"

"
 From the data available, there is little doubt that superimposed tablets communicate. Studies

based on"synchrotron microbeam X-ray diffraction (XRD) (Di Masi and Sarikaya, 2004), X-ray

photoelectron emission spectromicroscopy (X-PEEM) (Metzler et al., 2007; Gilbert et al., 2008)

and transmission electron microscopy (TEM) diffraction (SAED) (Gries et al., 2009b) show that

tablets stacked along a single column in the nacre of the abalone display similar contrast conditions,

i.e., they are co-oriented. Strict co-orientation of up to ten superimposed tablets in Mytilus edulis

has been demonstrated by means of electron backscatter diffraction (EBSD) (Dalbeck et al., 2006;

England et al., 2007). Our own data also support this view. Both in gastropods and cephalopods

nacre tablets stack in towers and the shapes of tablets of the same tower agree strictly (Fig. 1, A to

C). The same applies to the terraced nacre of bivalves, in which the shape of any growing tablet

always matches exactly that of the tablet immediately below (parental tablet) (Fig. 1, D and E).

Etched tablets of nacre show lineations, which we know coincide with the a-axis of the aragonite. In

the gastropod Gibbula plates sometimes display many sectors, with lineations intersecting at

approximately either 60 or 120 (Fig. 1F). There is a good match between the lineations of

superimposed tablets, although some sectors may appear or disappear from one layer to the next

(arrows in Fig. 1F). In bivalves, the lineations of superimposed tablets are coincident in all those

cases in which the parental-filial relation between them can be unequivocally established (Fig. 1G).

Electron-diffraction reveals an agreement in crystallographic orientation in the case of gastropods

(Fig. 1H) and cephalopods if the observed platelets belong to the same tower. In bivalves, an

agreement in orientation is often found between tablets of different lamellae in a zig-zag manner

due to the brickwall-like arrangement of tablets (Fig. 1I).

With some exceptions (Nudelman et al., 2006; Suzuki et al., 2009), authors have worked on

the assumption that MBs are a reality and have illustrated what may be MBs in scanning (SEM)

and/or TEM views (Fan and Yilong, 2001; Song et al., 2002; Su et al., 2002; Song and Bai, 2003;

Song et al., 2004; Rousseau et al., 2005; Berthelat et al., 2006; Velzquez-Castillo et al., 2006; Lin

4"

"
et al., 2008; Meyers et al., 2008; Fengzhang et al., 2009; Li and Huang, 2009; Saruwatari et al.,

2009; Espinosa et al., 2009). Width ranges of 36-54 nm (Song et al., 2004) and 25-55 nm (Gries et

al., 2009b) have been provided. A few authors have shown that the crystal lattice of the two

superimposed tablets is continuous across the MB (Gries et al., 2009b; Saruwatari et al., 2009).

TEM tomography in the abalone Haliotis laevigata has shown that what seem to be true mineral

connections are but the end of a continuum that initiates with simple hillocks (the so-called

nanoasperities, e.g., Berthelat et al., 2006) that protrude from both tablets and that do not connect

(Gries et al., 2009b). The intra-lamellar boundaries between tablets are densely studded with

purported MBs and nanoasperities (Gries et al., 2009b; Fan and Yilong, 2001; Song et al., 2002; Su

et al., 2002; Song and Bai, 2003; Song et al., 2004; Lin et al., 2008; Fengzhang et al., 2009; Li and

Huang, 2009; Saruwatari et al., 2009), such that a 3D estimate from their TEM or SEM views

implies that for a single tablet there would be hundreds of MBs. The only precise calculations (Song

et al., 2004) performed provide a figure of ~1600 MB per tablet.

On the other hand, wide connections (mean widths of 150-200 nm) have been described

between the tablets of the gastropods Gibbula and Monodonta located at the axes of nacre towers

(Checa et al., 2009). The interlamellar membrane is totally absent here and the crystal lattice is

continuous across them. The formation of the corresponding holes in the interlamellar membranes is

based on the recognized existence of a surface membrane that covers the mineralization

compartment of gastropod nacre and on its interaction with the incipiently growing tablets.

From the above summary, it is clear that there are two different conceptions as to what a MB

should be: the widespread opinion is that all minor connections between tablets are MBs, provided

that the crystal lattice is continuous across them; in the second view, MBs, at least in gastropods,

are wide connections centred on the tablets. Adherents to the former view have to deal with the

problem that the number of recognized MBs does not fit with the average one to one

correspondence between parental and filial tablets; in this case, it has to be assumed that most of the

5"

"
observed MBs do not progress. In view of the present imprecision regarding the nature of MBs, we

have examined the connections between tablets in representatives of the three main classes of

nacreous molluscs (gastropods, bivalves and cephalopods; the rare monoplacophorans are not

included owing to lack of material) by integrating data from different sources, mainly SEM and

TEM. Our goal is to determine which ones of those connections are likely candidates to be true

MBs, and so to present a unified picture of their formation in molluscs.

This paper is an extended version of the study we published last year (Checa et al, 2011).

The pioneering contributions of Wada and Nakahara

Hiroshi Nakahara, in his joint work with Bevelander (Bevelander and Nakahara, 1969) noted

for the first time that the interlamellar membranes are formed before the tablets. Later, in his 1979

work (Nakahara, 1979) he described the distribution of interlamellar membranes in gastropods,

together with the surface membrane. Moreover, in 1991, he described (Nakahara, 1991) the organic

core in the nacre towers of gastropods and gives his model of the formation of membranes and

tablets in bivalves and gastropods (Figs, 9 y 10, p. 346), which is that currently accepted.

In 1972 Koji Wada described a growth model (Wada, 1972; fig. 34, p. 158) that is clearly the

precursor of the mineral bridges of Schffer et al. (1997). The route from one tablet to another is via

pores in the membranes that he, along with Nakahara, had recognized. Although the study is not as

sophisticated as that of Schffer et al., a quarter of a century later, it is the authentic precursor of the

mineral-bridges idea, and must be recognized as such. The same scheme appears reproduced in

figure 10 of his work in Japanese of 1985 (Wada, 1985) and in fig. 3.19 of his book (Wada, 1999).

We dedicate this paper to Hiroshi Nakahara and Koji Wada whose pioneering work during

the 1960s-1980s, begun at the National Pearl Research Laboratory in Mie, Japan, established many

of the findings about nacre that our work has built upon, and whose contributions we sometimes

feel are not sufficiently recognized by the community.

6"

"
Materials and Methods

Shells of living specimens of the gastropods Gibbula umbilicalis, G. pennanti, Monodonta sp., and

Calliostoma zyzyphinus, the bivalve Nucula sulcata, and the cephalopod Nautilus pompilius were

fixed with 2.5% glutaraldehyde in a 0.1 M cacodylate buffer. Samples were CO2 critical point dried

for SEM observation. Other empty shells of the same species, as well as of the bivalves Acila

castrensis and Pinctada margaritifera were also observed. Some samples were etched with 10%

sodium hypochlorite for 30 mins, followed by 10 mins in a mixture of 25% glutaraldehyde and 10%

acetic acid. This is a variant of Mutveis protocol (details kindly provided by H. Mutvei,

Stockholm). Other samples were bleached with commercial bleach (4% active Cl) for several

minutes. In some polished sections the organic matrix was fixed with a mixture of 2.5%

glutaraldehyde and 2% formaldehyde; they were further demineralized with a methanolic solution

(3:1:6) in a gel medium (protocol by Hernndez-Hernndez et al., in prep.). All the treated shells

listed above were studied using a Leo Gemini 1530 field-emission SEM (FESEM) of the

Universidad de Granada.

For TEM, specimens of G. umbilicalis, P. margaritifera, N. pompilius, and A. castrensis were

first mechanically polished and subsequently thinned down to electron transparency with a GATAN

precision ion polishing system (PIPS) at the Fritz-Haber Institute of the Max-Planck Society in

Berlin. The 3-D arrangement of nacre was studied by slice-and-view focused ion beam (FIB)

tomography in the cases of the gastropod Osilinus lineatus and the bivalve P. margaritifera. Slicing

(at 20 nm) and viewing was performed with an Auriga Crossbeam (FIB-SEM) workstation

equipped with a Cobra ion column, belonging to Carl Zeiss Optronics GmbH (Oberkochen,

Germany). The 3-D reconstruction was performed with the program ImageJ (downloadable at

http://rsbweb.nih.gov/ij/). Two lamellae of P. margaritifera were also prepared for TEM with the

7"

"
Cobra ion column. TEM analysis was carried out using the Jeol 2200FS microscope at the

University of Aveiro, Portugal.

Results

Minor interruptions in the interlamellar membrane

Minor protuberances or outgrowths in the size range of 10-50 nm frequently interrupt the

interlamellar membranes in all species examined (Fig. 2, A to F). They are crystalline and generally

terminated by flat crystalline faces (Fig. 2, A to C and F). Often, their size is large enough to bridge

the gap between two tablets. In cases in which two superimposed tablets display an identical

diffraction contrast, high-resolution (HRTEM) lattice fringe imaging reveals that the crystal lattice

of tablets is continuous across such bridges (Fig. 2, A, B, D and E), a fact previously reported

(Gries et al., 2009b). However, intermediate domains with a slightly missoriented or distorted

lattice sometimes appear along the crystalline connections (Fig. 1, B, D and E).

STEM views (Fig. 2, C and F) demonstrate that this crystalline continuity is illusory in many

cases. The two-dimensional projection of a three-dimensional TEM lamella that contains hillock-

like protrusions which may overlap at different depths with respect to the viewing direction, gives

8"

"
Fig. 2. Minor vertical connections between tablets in gastropods and bivalves. (A-C) Gibbula umbilicalis. (D-

F) Pinctada margaritifera. A, B, D and E are high-resolution images of the interlamellar membranes. Despite

the presence of defects, there is apparent continuity of the crystal lattice across the contacts between

protrusions. C and F are STEM views of similar situations; it is clear that there is not true connection between

protrusions. (G) Bleached internal nacre of Nautilus pompilius. The nanocrystalline aspect is evident. (H-I)

Bleached nacre of Pinctada margaritifera. The surface is studded with nanocrystals, the aspect of which is

particularly euhedral in C.

9"

"
rise to the false impression of a continuous crystalline connection. STEM imaging also provides

cases where an intermediate gap, which otherwise goes unnoticed in the TEM view, becomes

clearly visible (Fig. 2, C, inset, and F). Some of the minor bridges in fact correspond to grains

(sometimes isolated) that have grown within the interlamellar membrane (Fig. 2C). SEM

observations of the surfaces of nacre tablets, i.e., the sites where minor connections may later

develop, reveal the presence of small protrusions (40-100 nm wide; Fig. 1,G and H). When viewed

from the top (Fig. 2I), the crystalline alignment of these nanocrystals becomes apparent.

Major interruptions of the interlamellar membrane

Gastropods and Nautilus margin.- Although it was formerly reported for gastropods (Checa et

al., 2009), we confirm here that in both cases, the major (> 100 nm) disruptions of the interlamellar

membrane appear aligned with the axes of the towers in TEM sections (Fig. 3, A and E). SEM

examination of some decalcified samples of gastropods shows that these breaks are circular and

appear strictly aligned with the tower axis (Fig. 3D). In TEM view, the membrane on both sides of

the interruptions is often curled towards the top of the tower (Fig. 3B). In Nautilus pompilius, the

interlamellar membrane is much less dense, with a frayed aspect and, sometimes, shows also

additional minor breaks (Fig. 3E) in the central region of a tower. As in bivalves, HRTEM imaging

and FFT analysis of the fringes demonstrate that there is crystal continuity between the two tablets

in contact and across the bridge in both groups (Fig. 3, C and F).

Bivalves.- TEM views of P. margaritifera, show occasional large (> 50 nm) interruptions.

When, moreover, the two tablets in contact show the same orientation, the crystal lattice is fully

continuous. The aspect of two of these instances in TEM and STEM views is shown in Fig. 3, G

and H. In these cases, the interlamellar membranes tend to be distorted close to the interruptions and

the gap is invariably larger than 100 nm. The material at the bridge shows exactly the same aspect

as in the bulk of the nacre tablets, including the typical presence of voids (Gries et al., 2009a).

10"

"
HRTEM and FFT imaging demonstrates that the bridge forms a crystallographic unit with the two

tablets in contact (Fig. 3, G and I).

Fig. 3. Mineral bridges in molluscs. (A-D) Gibbula umbilicalis. (E, F) Nautilus pompilius. (G-I) Pinctada

margaritifera. (A) TEM section through the axis of a tower displaying the interruptions of the interlamellar

membranes along the tower axis. (B) STEM view of two similar interruptions. They are adorned on each side

with upwards pointing twirls. (C) HRTEM view of the lower interruption in B. The crystal lattice is fully

continuous across the interruption and the FFT (inset) indicates that the tablets above and below and the

mineral bridge are one single crystal. (D) SEM view of a polished and decalcified sample. Note aligned holes

in the interlamellar membranes. (E) STEM view of a nacre tower showing aligned segments of the

11"

"
interlamellar membranes where they become interrupted and acquire a vanished aspect. The inset is a detail.

(F) TEM view of a mineral bridge. The HRTEM image (lower inset) and the FFTs (upper insets) imply

crystal continuity across the mineral bridge (framed area). (G-I) Two putative mineral bridges (G and H-I).

STEM views (lower insets in G and H) are also provided. The HRTEM images of the framed areas in G and

H (upper right insets in G and I) and the FFTs of the framed areas in G and I (upper left inset in G, and lower

right inset in I) imply both crystal unicity and continuity.

Slice and view of Osilinus and Pinctada

Osilinus lineatus.- The sliced towered nacre tablets show scattered minor connections across

the interlamellar membranes throughout as well as wide connections at their centres, which

delineate the tower axis (Fig. 4A). Around the connections there are particularly wide empty areas,

which in the 3D reconstruction of the interlamellar membranes (Fig. 4, A and B) appear as organic

necks around a central hole. Evaporation of the organic phase during the milling process may both

enlarge the organic neck and reduce the central hole.

Fig. 4. 3D reconstructions of the interlamellar membranes in the gastropod Osilinus lineatus (A-B) and in the

bivalve Pinctada margaritifera (C). (A) Plan view, the hole serving as passage for the mineral bridge is

marked with an arrow; other minor holes scattered on the membrane correspond to connecting nanoasperities.

(B) Side view, note organic neck around the hole. (C) Plan view; there are two putative passages indicated

with arrows. The broken lines mark the boundaries between tablets.

12"

"
Pinctada margaritifera.- Slicing and the associated 3D reconstruction of interlamellar membranes

(Fig. 4C) show that, besides many nm-sized connections there are occasional wide connections

between superimposed tablets. They have oval shapes, are several hundreds of nm wide (~ 150 x

400 nm in the examples shown in Fig. 4C), and occur preferentially at the lateral boundaries

between tablets.

Connections between tablets in gastropods and Nautilus

The number of minor connections between two superimposed tablets that strictly coincide in

orientation, even if we consider only those that fulfil the continuous-lattice condition, would be at

least several tens, a too-high figure taking into account that each parental tablet originates on

average only one tablet. Even if tablets are polycrystalline, the number of sectors of the filial tablet

may vary within narrow limits, but remains statistically constant. This implies that the attribution of

the minor connections as true MBs is more than doubtful. This fact, together with the continuous

gradation from unconnected hillocks, to hillocks that overlap in the third dimension and are

connected only in appearance, to fully connected (apparent MBs), imply that these last cases are

obtained when hillocks from both tablets protrude and adapt so tightly that the resulting feature

appears to be continuous. This is evident in STEM view (Fig. 2C). The so-called hillocks or

nanoasperities are merely nanocrystalline outgrowths or nanocrystals growing on the surfaces of the

platelets, of the kind shown in Fig. 2, G to I. The latter probably grow within the proteinaceous part

of the interlamellar membranes, where they have space enough to develop crystalline faces.

Major connections were recognized and described in gastropods (Checa et al., 2009). We

have here confirmed their nature and alignment not only in gastropods, but also in the towered

nacre of the cephalopod Nautilus. In this group, the interlamellar membrane may either be lacking,

or, alternatively, present a frayed aspect, as if it had had insufficient time to organize completely.

The nature of major connections as true MBs also fits in with their position at the very centre of the

13"

"
tablets. The model for their origin in gastropods is based on the fact that the tips of the early nuclei

of tablets grow partly embedded within the relatively thick (~100 nm) surface membrane (Checa et

al., 2009) (Fig. 5A). This prevents the interlamellar membrane from extending through these sites,

and hence there is crystal continuity. SEM observations reveal the presence of a similar membrane

in immature specimens of Nautilus pompilius (Fig. 5, B and C). As in gastropods, it covers the

nacre compartment (Fig. 5, B, inset, and C). Its thickness in SEM view is some 100-200 nm; we

still lack details as to its ultrastructure and relationship to the interlamellar membranes.

Fig. 5. Surface membrane in gastropods and Nautilus. (A) Fracture (side) view of the nacre compartment of

the gastropod Gibbula pennanti, which is covered by the vesiculate surface membrane. (B) Surface view of

the nacre close to the aperture of Nautilus pompilius. The surface membrane has cracked upon dehydration.

The inset shows one fracture, which allows us to see the interlamellar membranes below the surface

membrane. (C) Side view of the marginal nacre of Nautilus pompilius. Upon dehydration, the surface

membrane has collapsed and adhered to the towers of nacre. Note the difference in thickness between the

surface membrane and the interlamellar membranes. sm = surface membrane; ilm = interlamellar membrane.

The only difference between gastropods and Nautilus is that interlamellar membranes are not

totally absent at the tower axis in the latter. Indeed, they are strongly perforated in that region.

Nevertheless, the similarity in tablet distribution and the existence of a surface membrane analogous

to that in gastropods (Fig. 5, B and C) imply that the mode of formation of mineral bridges in

14"

"
Nautilus is similar to that in gastropods, although the relationship of the surface membrane to the

nascent tablets and interlamellar membranes must differ.

Connections between tablets in bivalves

As in the former groups, minor and major connections have been recognized, with minor

connections having exactly the same aspect and distribution. Each bivalve nacre tablet has from

zero to two, very rarely three, new filial tablets. The same reasons as to the aspect, density and

distribution of minor connections adduced in gastropods and Nautilus to discard them as real MBs

apply here.

Despite their elusive nature, major interruptions in the interlamellar membrane of bivalves

have been recognized in the 3D reconstructions and observed directly with TEM. The aspect of the

material through the MB is exactly the same as in the tablet interior and the crystal lattice is fully

continuous across the MB, without any indication of crystalline defects. The membrane has the

aspect of having been disrupted at the place where the bridge formed. Such interruptions are perfect

candidates for being crystalline connections between tablets, particularly since they appear close to

the margins of the parental tablets and at the centres of the filial tablets, where, the continuation

from below serves as a core for a newly forming tablet, thus explaining exactly what is observed

with SEM and TEM (Fig. 1, E and I).

Any mineral-bridge-based explanation for the genesis of bivalve tablets has to take into

account this fact that filial tablets grow from the edges of the parent tablet below, i.e., pores in the

interlamellar membrane occur most often near the borders of tablets, exactly as in our 3D

reconstruction. This implies that the production of new tablets happens when the parental tablet is at

an advanced stage of growth or, in other words, when the volume of not yet mineralized material

enclosed by the growing tablets of one layer shrinks. We hypothesize that in these cases pore

formation takes place by rupture of the interlamellar membrane owing to osmotic pressure. There

15"

"
begins to be an osmotic pressure difference across an interlamellar membrane when tablet growth

closes the compartment below the membrane, which acts as a lid on the compartment, while the

zone above the membrane remains open to the general extrapallial space (Fig. 6, A and B). The

composition of the extrapallial liquid in the lidded compartment alters as calcium carbonate

crystallization proceeds with organic compounds becoming ever more concentrated as compartment

volume diminishes. The membrane is almost certainly semipermeable, with small molecules such as

water passing across it, while the larger molecules are unable to cross. Thus the increase in the

concentration of organic compounds implies a corresponding increase in the osmotic pressure

exerted on the lower side of the interlamellar membrane. This imbalance across the membrane leads

to its bowing upwards towards the edge of tablets (Fig. 6B1), as is noted in slices through bivalve

nacre (Fig. 7), and eventually to its rupture at a weak point (Fig. 6B2). Immediately after the

osmotic pressure is released, the burst membrane bends into the open void, and a valley is usually

observed at the very junction between tablets in areas supposedly close to the mineral bridges (Fig.

Fig. 6. Model for the formation of mineral bridges in bivalves. (A, B) Neighbouring tablets grow until a

closed space, i.e., one with no connection to the rest of the interlamellar space, is formed. (B1-B3) Close up

view of the processes occurring in this space. Increasing concentration of organic molecules raises the

osmotic pressure within the lidded compartment and causes the interlamellar membrane to bow upwards (B1),

16"

"
until a break is produced (B2). Once the osmotic pressure is released, the interlamellar membrane relaxes and

a new tablet emerges from the edge of the tablet which first reaches the tear (B3).

6B3, and Fig. 7, arrows). This mechanism explains why pores form preferentially at the edges of

tablets and also fits in with the elongated shape of the reconstructed major pores in the membranes

(Fig. 4C).

Fig. 7. SEM view of the polished nacre of Pinctada margaritifera. The interlamellar membranes are convex

upwards close to the lateral contacts between tablets and the incised small valley at the contacts (arrows).

Growth direction is to the top.

Discussion

Our study, together with former evidence (Di Masi and Sarikaya, 2004; Metzler et al., 2007; Gilbert

et al., 2008; Gries et al., 2009b, Dalbeck et al., 2006; England et al., 2007) leads us to conclude that

there is an exact match in orientation between a given tablet and the one(s) that grow(s) on top of it.

This feature evidently calls for the necessary communications of aragonite crystals across the

interlamellar membrane.

Nacre tablets in gastropods (Checa et al., 2009) and Nautilus communicate across circular

interruptions in the interlamellar membrane, which are persistently found aligned along the axes of

17"

"
the towers. In both groups, the formation of axial interruptions is related to the existence of a

surface membrane, although the details as to the formation of MBs may differ slightly between the

groups. In bivalves, interruptions are placed close to the edges of the parental tablets. In this case,

we assume that their origin is secondary, due to differences in osmotic pressure, which we

hypothesize to cause bursting of the interlamellar membrane in the last stages of growth of the

underlying tablets. Differences in osmotic pressure across the membrane in gastropods, if present,

are so in an attenuated manner. Firstly, both the horizontal and vertical gradients in tablet growth

are much lower in gastropods than in bivalves. In the latter group, fully grown tablets change into

nuclei in a single lamella over a distance of several tens of microns, whereas in gastropods, tablets

in a given lamella are in a similar growth state. In the vertical direction, initial nuclei of bivalves sit

on top of fully-grown or almost-fully-grown tablets; this is not the case in gastropods, particularly

when towers have a high aspect ratio. Finally, as the membranes of gastropods appear porous

(Schffer et al., 1997; Cartwright and Checa, 2007), while those of bivalves do not (Cartwright and

Checa, 2007; Cartwright et al., 2009), we hypothesize that the former may be transparent to organic

molecules. In any case, the number and distribution of the described major breaks exactly match

those of filial tablets that emerge once a new lamella is being formed. This is not the case with other

described minor breaks (<50 nm), which have been considered to be true MBs (Gries et al., 2009b).

On this basis, as well as on that of their morphology and distribution, we disregard them as such.

Our model implies that in nacre there is full continuity of aragonite crystals across the

interruptions in the membranes; any kind of heteroepitaxy being therefore excluded. Hence, since

there is no need for a nucleation event every time a new tablet is to be formed, we consider the

models based on the recognition of aragonite nucleating proteins and their role in the determination

of the crystallographic orientations (Suzuki et al.; 2009, Krger, 2009) as superfluous.

Our study in no way exhausts the issue of how nacre grows, but rather aims at providing a

new perspective for future studies. For example, there can be little hope for any effort aiming to

18"

"
mimic nacre in laboratory that does not take into account the increasing body of knowledge on this

fascinating natural product.

Acknowledgements

We acknowledge Achim Klein (Fritz-Haber Institut, Berlin) and Fabian Prez-Willard (Zeiss Optronics

GmbH, Oberkochen, Germany) for sample preparation and analysis. Funding was provided by Projects

CGL2007-60549, CGL2010-20748-CO2-01 and FIS2010-22322-C02-02 of the Spanish Ministerio de Ciencia

e Innovacin, Projects REDE/1509/RME/2005 and PTDC/CTM/100468/2008 of the Portuguese Foundation

for Science and Technology (FCT), the Research Group RNM190 (Junta de Andaluca), and the European

COST Action TD0903.

References

Berthelat, F., Li, C. M., Comi, C. and Espinosa, H. D. (2006): Mechanical properties of nacre constituents and

their impact on mechanical performance. J. Mater. Res., 21, 1977-1986.

Bevelander, G. and Nakahara, H. (1969): An electron microscope study of the formation of the nacreous layer

in the shell of certain bivalve molluscs. Calcif. Tissue Res, 3, 84-92.

Cartwright, J. H. E. and Checa, A. G. (2007): The dynamics of nacre self-assembly. J. R. Soc. Interface, 4,

491-504.

Cartwright, J. H. E., Checa, A., Escribano, B. and Sainz-Daz, C. I. (2009): Spiral and target patterns in

bivalve nacre manifest a natural excitable medium from layer growth of a biological liquid crystal. Proc.

Natl. Acad. Sci. U.S.A., 106, 10499-10504.

Checa, A. G., Cartwright, J. H. E. and Willinger, M.-G. (2009): The key role of the surface membrane in why

gastropod nacre grows in towers. Proc. Natl. Acad. Sci. USA, 106, 38-43.

Checa, A. G., Cartwright, J. H. E. and Willinger, M.-G. (2011): Mineral bridges in nacre. J. Struct. Biol., 176,

330-339.

Dalbeck, P., England, J., Cusack, M., Lee, M. R. and Fallick, A. E. (2006): Crystallography and chemistry of

the calcium carbonate polymorph switch in M. edulis shells. Eur. J. Mineral., 18, 601-609.

19"

"
Di Masi, E. and Sarikaya, M. (2004): Synchrotron X-ray microbeam diffraction from abalone shell. J. Mater.

Res., 19, 1471-1476.

England, J., Cusack, M., Dalbeck, P. and Prez-Huerta, A. (2007): Comparison of the crystallographic

structure of semi nacre and nacre by electron backscatter diffraction. Crystal Growth Des., 7, 307-310.

Espinosa, H. D., Rim, J. E,. Barthelat, F. and Buehler, M. (2009): Merger of structure and material in nacre

and bone Perspectives on de novo biomimetic materials. Prog. Mater. Sci., 54, 1059-1100.

Fan, S. and Yilong B. (2001): Mineral bridges in nacre and its effects. Acta Mech. Sinica (Engl. Ser.), 17,

251-257.

Fengzhang, R., Xindi, M., Zhanhong, M. and Juanhua, S. (2009): Study on microstructure and

thermodynamics of nacre in mussel shell. Mater. Chem. Phys., 114, 367-370.

Gilbert, P. U. P. A., Metzler, R. A., Zhou, D., Scholl, A., Doran, A., Young, A., Kunz, M., Tamura N. and

Coppersmith, S. N. (2008). Gradual ordering in red abalone nacre. J. Am. Chem. Soc., 130, 17519-

17527.

Gries, K., Krger, R., Kbel, C., Fritz, M. and Rosenauer, A. (2009a): Investigations of voids in the aragonite

platelets of nacre. Acta Biomater., 5, 3038-3044.

Gries, K., Krger, R., Kbel, C., Schowalter, M., Fritz, M. and Rosenauer, A. (2009b): Correlation of the

orientation of stacked aragonite platelets in nacre and their connection via mineral bridges.

Ultramicroscopy, 109, 230-236.

Krger, N. (2009): The molecular basis of nacre formation. Science, 325, 1351-1352.

Levi-Kalisman, Y., Falini, G., Addadi, L. and Weiner, S. J. (2001): Structure of the nacreous organic matrix

of a bivalve mollusc shell examined in the hydrated state using cryo-TEM. J. Struct. Biol., 135, 8-17.

Li, X. and Huang, Z. (2009): Unveiling the formation mechanism of pseudo-single-crystal aragonite platelets

in nacre. Phys. Rev. Lett., 102, 075502.

Lin, A. Y.-M. Chen, P.-Y. and Meyers, M. A. (2008): The growth of nacre in the abalone shell. Acta

Biomater., 4, 131-138.

Metzler, R. A., Abrecht, M., Olabisi, R. M., Ariosa, D., Christopher, J. J., Frazer, B. H. Coppersmith, S. N.

and Gilbert, P. U. P. A. (2007): Architecture of columnar nacre, and implications for its formation

mechanism. Phys. Rev. Lett., 98, 26810.

20"

"
Meyers, M. A., Lin, A. Y.-M., Chen, P. Y. and Muyco, J. J. (2008): Mechanical strength of abalone nacre:

Role of the soft organic layer. Mech. Behav. Biomed. Mat., 1, 76-85.

Nakahara, H. (1979): An electron microscope study of the growing surface of nacre in two gastropod species,

Turbo cornutus and Tegula pfeifferi. Venus, 3, 205-211.

Nakahara, H. (1983): Calcification of gastropod nacre. p. 225230. In Biomineralization and Biological Metal

Accumulation, ed. by P. Westbroek and E. W. De Jong, Reidel, Dordrecht.

Nakahara, H. (1991): Nacre formation in bivalve and gastropod mollusks. p. 343350. In Mechanisms and

Phylogeny of Mineralization in Biological Systems, ed. By S. Suga and H. Nakahara, Springer, Tokyo.

Nudelman, F., Gotliv, B. A., Addadi, L. and Weiner, S. J. (2006): Mollusk shell formation: Mapping the

distribution of organic matrix components underlying a single aragonitic tablet in nacre. J. Struct. Biol.,

153, 176-187.

Rousseau, M., Lopez, E., Stempfl, P., Brendl, M., Franke, L., Guette, A., Naslain, R. and Bourrat, X.

(2005): Multiscale structure of sheet nacre. Biomaterials, 26, 6254-6262.

Saruwatari, K., Matsui, T., Mukai, H., Nagasawa, H. and Kogure, T. (2009): Nucleation and growth of

aragonite crystals at the growth front of nacres in pearl oyster, Pinctada fucata. Biomaterials, 30, 3028-

3034.

Schffer, T. E., Ionescu-Zanetti, C., Proksch, R., Fritz, M., Walters, D. A., Almquist, N., Zaremba, C. M.,

Belcher, A. M., Smith, B. L., Stucky, G. D., Morse, D. E. and Hansma, P. K. (1997): Does abalone nacre

form by heteroepitaxial nucleation or by growth through mineral bridges? Chem. Mater., 9, 1731-1740.

Song, F. and Bai, Y. L. J. (2003): Effects of nanostructures on the fracture strength of the interfaces in nacre.

J. Mater. Res., 18, 1741-1744.

Song, F., Soh, A. K. and Bai, Y. L. (2004): Structural and mechanical properties of the organic matrix layers

of nacre. Biomaterials, 24, 3623-3631.

Song, F., Zhang, X. H. and Bai, Y. L. (2002): Microstructure and characteristics in the organic matrix layers

of nacre. J. Mater. Res., 17, 1567-1570.

Su, W., Belcher, A. M., Zaremba, C. M., Morse, D. E., Stucky, G. D. and Heuer, A. H. (2002): Structural and

microstructural characterization of the growth lines and prismatic microarchitecture in red abalone shell

and the microstructures of abalone flat pearls. Chem. Mater., 14, 3106-3117.
21"

"
Suzuki, M., Saruwatari, K., Kogure, T., Yamashita, Y., Nishimura, T., Kato, T. and Nagasawa, H. (2009): An

acidic matrix protein, Pif, is a key macromolecule for nacre formation. Science, 325, 1388-1390.

Velzquez-Castillo, R. R., Reyes-Gasga, J., Garca-Gutirrez, D. I. and Jos-Yacamn, M. (2006): Crystal

structure characterization of Nautilus shell at different length scales. Biomaterials 27, 45084517.

Wada, K. (1972): Nucleation and growth of aragonite crystals in the nacre of some bivalve molluscs.

Biomineralization Res. Rep., 6, 141-159.

Wada, K. (1985): Growth of CaCO3 crystals in bivalve shell mineralization, Journal of the Japanese

Association of Crystal Growth, vol. 12, no 1 & 2, 57-70 (in Japanese with English abstract).

Wada, K. (1999) Shinju no Kagaku [The science of pearls], Shinju Shinbun, Tokyo (in Japanese).

Weiner, S. and Traub, W. (1980): X-ray diffraction study of the insoluble organic matrix of mollusk shells.

FEBS Lett., 111, 311-316.

Weiner, S. Traub, W. (1981): Structural aspects of recognition and assembly in biological macromolecules. p.

467482. In Biological Macromolecules, ed. by M. Balaban, J. L. Sussman, W. Traub and A. Yonath,

Balaban ISS, Rehovot, and Philadelphia.

22"

"